SDI² — Software-Defined Intelligent Intersections Hybrid SIMP–AIMP Intelligent Traffic Signal Management Using SUMO & SDN Principles
This repository contains the full implementation and SUMO simulation dataset for SDI² (Software-Defined Intelligent Intersections), a hybrid traffic management framework that switches between:
SIMP — Synchronous Intersection Management Protocol
AIMP — Adaptive Intersection Management Protocol
depending on real-time vehicle density and grouping.
This project is based on the EWGT 2025 research work titled: “Software-Defined Intelligent Intersections for Smart Mobility”
📂 Project Structure
📦 final result/
├── pycache/
│ ├── trafficmetrics.cpython-312.pyc
│ └── trafficsignalcontroller.cpython-312.pyc
├── data/ │ ├── AIMP/ # Results/snapshots for AIMP evaluation │ ├── cross.add # SUMO additional elements │ ├── cross.det # Detector definitions │ ├── cross.edg # Network edge definitions │ ├── cross.flow # Traffic flow definitions │ ├── cross.net # Road network │ ├── cross.netccfg # NETCONVERT config │ ├── cross.nod # Nodes (intersections) │ ├── cross.out # Output files │ ├── cross.rou # Route files │ ├── cross.src # Source lane config │ ├── e0_0, e0_1, e1_0 ... # Network state snapshots │ ├── SUMO Configuration File # Simulation config (.sumocfg) │ └──
├── output/ │ ├── 0.133 # Simulation output logs │ └── tripinfo # SUMO trip-level performance metrics
└── result.py # Main Python script (SIMP/AIMP decision logic)
🚀 SDI² Overview
SDI² is an SDN-inspired traffic control system that:
✔ Reads real-time traffic data from SUMO ✔ Detects isolated vs. grouped vehicles ✔ Dynamically switches between SIMP and AIMP ✔ Adjusts traffic signal phases automatically ✔ Computes performance metrics (stopped delay, emissions, fuel use) 🧠 Core Components
- SIMP — Synchronous Intersection Management Protocol
Used when:
A single isolated vehicle is approaching the intersection.
Features:
Pre-defined phase timings
Conflict-free directions (CDM-based scheduling)
One vehicle per non-conflicting lane
- AIMP — Adaptive Intersection Management Protocol
Used when:
Multiple vehicles intend to cross in the same direction
Features:
Dynamic adjustment of green duration
Batch/group serving
Improved throughput
Reduced idle delay
- SDI² Mode Switching Logic (Inside result.py)
Pseudo-logic reflecting your implementation:
if consecutive_vehicle_count(direction) > 1: activate_AIMP() else: activate_SIMP()
🧪 SUMO Simulation Setup (data folder)
data/ folder contains the complete network:
File Purpose cross.net, cross.nod, cross.edg Road network & node geometry cross.flow Traffic injection definition cross.det Induction loop detectors cross.rou Vehicle routes cross.src Source lane mapping cross.out, e0_0, etc. Output and network state files *.sumocfg Main SUMO simulation config file
This structure represents an 8-inflow, 4-arm intersection used in the EWGT evaluation.
📊 Simulation Output (output folder)
The output/ directory stores:
tripinfo → Per-vehicle travel time, delay, stops
0.133 → Aggregated emission/fuel metrics
These files are used to compute:
Stopped delay
Fuel consumption
PMx emissions
As reported in the EWGT 2025 paper.
📈 Performance Summary (from your abstract)
(Backed by SUMO results from the project) SDI² achieves:
Metric Improvement vs RR Improvement vs SIMP Stopped Delay 89.5% ↓ 5% ↓ Fuel Consumption 63.3% ↓ 15% ↓ PMx Emissions 76.9% ↓ 27% ↓
This is due to efficient SIMP–AIMP switching.
🏃♂️ How to Run the Project
- Install SUMO
Download from: https://sumo.dlr.de/docs/Downloads.html
Ensure SUMO commands are available:
sumo --version
-
Install Python Dependencies pip install sumolib traci
-
Run the Simulation
Inside final result/:
python result.py
This script:
Loads the SUMO network from data/
Runs the simulation
Applies SDI² logic
Stores outputs in output/
🧩 What result.py Does
traci connection setup
sensor data extraction
SIMP logic
AIMP phase adjustment
Decision switching
Metrics calculation (linked to trafficmetrics.py)
🎯 Future Enhancements
Multi-intersection SDN coordination
Neural-network–based phase prediction
V2I communication integration
Real-world RSU integration
👤 Developer
CHEPURI DILEEP Developer & Implementer of the SDI² Simulation Framework Department of AI, Sree Vidyanikethan Engineering College, India
📜 License
MIT or CC-BY-NC-ND (based on paper)